[prev in list] [next in list] [prev in thread] [next in thread]
List: r-sig-finance
Subject: Re: [R-SIG-Finance] Cointegration question.
From: Matt Rimmer <rimmer.matt () gmail ! com>
Date: 2013-06-15 20:19:11
Message-ID: 6D51B040-F6C2-45DA-B7E8-70FFAFB4D522 () gmail ! com
[Download RAW message or body]
Ganesha/All,
If using a long/short pairs strategy is your ultimate goal, in addition to Kent's \
comments below and from my own interest and research, you may want to consider that, \
whatever timeframe you're looking at, there are other tools used by professionals to \
build paired strategies other than a cointegration test.
Some recent examples include;
Intraday:
http://dspace.mit.edu/bitstream/handle/1721.1/59122/658860705.pdf
Longer-term::
http://turnkeyanalyst.com/2013/06/pairs-trading-kicks-butt/?utm_source=empiricalfinanc \
e%2Fblog&utm_medium=Empirical+Finance+Blog&utm_campaign=Feed%3A+turnkeyanalyst+%28Turnkey+Analyst+Blog%29
Matt
On 15 Jun 2013, at 15:15, Kent Hoxsey <khoxsey@gmail.com> wrote:
If you are looking for tools to identify trading pairs, you might start with Bryan \
Lewis' presentation from R-Finance 2012:
http://www.rinfinance.com/agenda/2012/talk/BryanLewis.pdf
GL!
On Jun 15, 2013, at 5:30 AM, ganesha0701 wrote:
> Thanks for the inputs Marc. You provide some interesting insights. Yes,
> they are all prices. In my case the prices are very close to difference
> stationary actually. But none the less, any particular tool that you will
> recommend when it comes to testing for pairs trading. That is the ultimate
> application of interest, which is why I was interested in cointegration in
> the first place.
>
>
> On Sat, Jun 15, 2013 at 1:54 AM, Wildi Marc (wlmr) <wlmr@zhaw.ch> wrote:
>
> > Ganesha, Brian, All
> >
> > The log-return transformation typically eliminates trends of `prices' (the
> > latter should behave not too far away from a random-walk although we all
> > know that's not entirely true because otherwise this Mailing list wouldn't
> > exist). Therefore the empirical significance Level of the ADF-test should
> > be markedly below 5% for log-Returns (except if there is/are shift(s) in
> > the transformed data!). The posted results (25%) strongly suggest Prices
> > (not log-Returns).
> >
> > Cointegration: this is an econometrician tool developped for `stable'
> > (difference-stationary Gaussian) series which `behave well' over longer
> > time spans: Forget about application of this very sensitive stuff to
> > non-stationary financial data. Prices are not difference-stationary!
> > Econometrician are interested in the DGP (data generating process), not in
> > generating trading performances: therefore typical optimization criteria
> > are misleading: all statistics address one-step ahead mean-square
> > performances; who in the world (besides econometrician) is interested in
> > such a criterion?
> >
> > My advice: skip this unreliable Topic and save some time for leisure!
> >
> > Marc
> >
> > ________________________________________
> > Von: r-sig-finance-bounces@r-project.org [
> > r-sig-finance-bounces@r-project.org]" im Auftrag von "Brian G.
> > Peterson [brian@braverock.com]
> > Gesendet: Freitag, 14. Juni 2013 18:14
> > An: r-sig-finance@r-project.org
> > Betreff: Re: [R-SIG-Finance] Cointegration question.
> >
> > Please don't repost. If someone has the answer to your question and
> > feels like helping, they will.
> >
> > The most common problem we see in the list archives when questions like
> > this arise is that people are trying to test stationarity and
> > cointegration on prices rather than on returns.
> >
> > However, you haven't actually provided reproducible data with your
> > partial code, so without that I'm just guessing.
> >
> > - Brian
> >
> > On 06/14/2013 11:09 AM, ganesha0701 wrote:
> > > I have two time series that I am investigating, acc and amb, the time
> > > frequency is daily data. They are both non stationary, as evidenced by
> > the
> > > follows.
> > >
> > >
> > >
> > > adf.test(df$acc)
> > >
> > > Augmented Dickey-Fuller Test
> > >
> > > data: df$acc
> > > Dickey-Fuller = -2.7741, Lag order = 5, p-value = 0.2519
> > > alternative hypothesis: stationary
> > >
> > > > adf.test(df$amb)
> > >
> > > Augmented Dickey-Fuller Test
> > >
> > > data: df$amb
> > > Dickey-Fuller = -1.9339, Lag order = 5, p-value = 0.6038
> > > alternative hypothesis: stationary
> > >
> > > I am looking to test for cointegration between the two time series but
> > the
> > > problem I am running into is that the cointegrating vector seems to
> > change
> > > in time.
> > >
> > >
> > > 1)* First 200 points*
> > >
> > > ######################
> > > # Johansen-Procedure #
> > > ######################
> > >
> > > Test type: maximal eigenvalue statistic (lambda max) , with linear trend
> > >
> > > Eigenvalues (lambda):
> > > [1] 0.0501585398 0.0003129906
> > >
> > > Values of teststatistic and critical values of test:
> > >
> > > test 10pct 5pct 1pct
> > > r <= 1 | 0.06 6.50 8.18 11.65
> > > r = 0 | 10.19 12.91 14.90 19.19
> > >
> > > Eigenvectors, normalised to first column:
> > > (These are the cointegration relations)
> > >
> > > acc.l2 amb.l2
> > > acc.l2 1.0000000 1.000000
> > > amb.l2 -0.9610573 -2.237141
> > >
> > > Weights W:
> > > (This is the loading matrix)
> > >
> > > acc.l2 amb.l2
> > > acc.d -0.03332428 -0.002576070
> > > amb.d 0.03986111 -0.001591227
> > >
> > >
> > > 2) *First 1000 points*
> > >
> > > ######################
> > > # Johansen-Procedure #
> > > ######################
> > >
> > > Test type: maximal eigenvalue statistic (lambda max) , with linear trend
> > >
> > > Eigenvalues (lambda):
> > > [1] 0.019211132 0.001959403
> > >
> > > Values of teststatistic and critical values of test:
> > >
> > > test 10pct 5pct 1pct
> > > r <= 1 | 1.96 6.50 8.18 11.65
> > > r = 0 | 19.36 12.91 14.90 19.19
> > >
> > > Eigenvectors, normalised to first column:
> > > (These are the cointegration relations)
> > >
> > > acc.l2 amb.l2
> > > acc.l2 1.0000000 1.00000
> > > amb.l2 -0.8611314 15.76683
> > >
> > > Weights W:
> > > (This is the loading matrix)
> > >
> > > acc.l2 amb.l2
> > > acc.d -0.008993595 -0.0002419353
> > > amb.d 0.027935684 -0.0002067523
> > >
> > >
> > > 3)* Whole History*
> > >
> > > ######################
> > > # Johansen-Procedure #
> > > ######################
> > >
> > > Test type: maximal eigenvalue statistic (lambda max) , with linear trend
> > >
> > > Eigenvalues (lambda):
> > > [1] 0.0144066813 0.0008146258
> > >
> > > Values of teststatistic and critical values of test:
> > >
> > > test 10pct 5pct 1pct
> > > r <= 1 | 1.16 6.50 8.18 11.65
> > > r = 0 | 20.64 12.91 14.90 19.19
> > >
> > > Eigenvectors, normalised to first column:
> > > (These are the cointegration relations)
> > >
> > > acc.l2 amb.l2
> > > acc.l2 1.0000000 1.00000
> > > amb.l2 -0.8051537 -25.42806
> > >
> > > Weights W:
> > > (This is the loading matrix)
> > >
> > > acc.l2 amb.l2
> > > acc.d -0.01003068 7.009487e-05
> > > amb.d 0.02128464 6.980209e-05
> > >
> > > You can see the marginal change the coefficient values, from -0.96 to
> > -0.86
> > > to -0.80.
> > >
> > > My question is how to interpret this, what is the optimal look back
> > period,
> > > what is the true relationship I should use for future prediction?
> >
> > _______________________________________________
> > R-SIG-Finance@r-project.org mailing list
> > https://stat.ethz.ch/mailman/listinfo/r-sig-finance
> > -- Subscriber-posting only. If you want to post, subscribe first.
> > -- Also note that this is not the r-help list where general R questions
> > should go.
> > _______________________________________________
> > R-SIG-Finance@r-project.org mailing list
> > https://stat.ethz.ch/mailman/listinfo/r-sig-finance
> > -- Subscriber-posting only. If you want to post, subscribe first.
> > -- Also note that this is not the r-help list where general R questions
> > should go.
>
> [[alternative HTML version deleted]]
>
> _______________________________________________
> R-SIG-Finance@r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-finance
> -- Subscriber-posting only. If you want to post, subscribe first.
> -- Also note that this is not the r-help list where general R questions should go.
_______________________________________________
R-SIG-Finance@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-sig-finance
-- Subscriber-posting only. If you want to post, subscribe first.
-- Also note that this is not the r-help list where general R questions should go.
[[alternative HTML version deleted]]
_______________________________________________
R-SIG-Finance@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-sig-finance
-- Subscriber-posting only. If you want to post, subscribe first.
-- Also note that this is not the r-help list where general R questions should go.
[prev in list] [next in list] [prev in thread] [next in thread]
Configure |
About |
News |
Add a list |
Sponsored by KoreLogic